Cathodic generation and oxidation of lead zintl ion Pb94- in potassium iodide solutions in liquid ammonia

Article abstract of DOI:10.1021/jp9630737

The electrochemical reduction of lead and oxidation reaction of the resulting Pb₉^4- Zintl ions have been studied in 0.1 and 1.0 M KI solutions in liquid ammonia at -70°C. The reduction process is a complicated multistep reaction. Under certain conditions the oxidation of Pb₉^4- occurs as a simple, electrochemically irreversible reaction for which the limiting current is diffusion controlled. The diffusion coefficient of Pb₉^4- was determined from a combination of chronoamperometry measurements on a 25 μm Pt ultramicroelectrode and chronocoulometry measurements on a 1 mm Pb disk electrode. The chronocoulometric method for the determination of Pb₉^4- determination is presented and used to determine the actual Pb₉^4- concentrations in the electrochemical cell. Side effects, which accompany the reduction and oxidation reactions, have also been studied and the ways to avoid them presented.

Full text of DOI:10.1021/jp9630737

J. Phys. Chem. B 1997, 101, 2867-2873
4-
2867
Cathodic Generation and Oxidation of Lead Zintl Ion Pb9 in Potassium Iodide Solutions
in Liquid Ammonia
Jerzy B. Chlistunoff and Joseph J. Lagowski*
Department of Chemistry and Biochemistry, UniVersity of Texas in Austin, Austin, Texas 78712
ReceiVed: October 7, 1996; In Final Form: January 31, 1997X
4-
The electrochemical reduction of lead and oxidation reaction of the resulting Pb9 Zintl ions have been
studied in 0.1 and 1.0 M KI solutions in liquid ammonia at -70 °C. The reduction process is a complicated
4-
multistep reaction. Under certain conditions the oxidation of Pb9 occurs as a simple, electrochemically
4-
irreversible reaction for which the limiting current is diffusion controlled. The diffusion coefficient of Pb9
was determined from a combination of chronoamperometry measurements on a 25 µm Pt ultramicroelectrode
and chronocoulometry measurements on a 1 mm Pb disk electrode. The chronocoulometric method for the
4-
4-
determination of Pb9 determination is presented and used to determine the actual Pb9 concentrations in
the electrochemical cell. Side effects, which accompany the reduction and oxidation reactions, have also
been studied and the ways to avoid them presented.
Introduction
suitable for electrochemical measurements than ethylenediamine,
which has the permittivity of 12.9.
The first observation of the species we call today Zintl ions1
was reported more than 100 years ago.2,3 These homopoly-
atomic anions of metals (metaloids) of group IV and V are
formed by dissolution of alloys of these elements with alkali
metals in liquid ammonia or ethylenediamine or by a direct
reaction between the elements in one of the above solvents.1
The first successful attempt to generate Zintl ions electrochemi-
cally dates back to the beginning of this century.4 Since the
initial studies,2-12 our understanding of Zintl ions has signifi-
Experimental Section
A three-compartment electrochemical cell equipped with
medium-porosity glass frits was used in the experiments.
A lead disk electrode was prepared by sealing of a 1 mm
lead wire (Puratronic, Alfa Aesar, Ward Hill, MA) in an epoxy
resin (Huntington Mechanical Laboratories Inc., Mountain View,
CA) or shrinkable PVC tubing.
cantly improved. Also, the name “Zintl ions” has been extended
Large lead electrodes used for preparative electrolysis were
prepared by melting lead wire and pouring the liquid into a
porcelain boat. A 1.5 mm Pt wire was heated in a flame until
it started melting and the hot end pressed against the lead bar
which locally melted and formed a durable electrical contact.
The large lead electrode was usually around 3 cm long and has
a diameter of approximately 7 mm. During experiments around
two-thirds of its length was immersed in the solution. Platinum
disk electrodes (0.5 mm diameter) as well as Pt ultramicroelec-
trodes (25 µm diameter) were prepared by sealing corresponding
Pt wires in a soft glass. In most experiments a silver wire (0.5
mm diameter) was used as a quasireference electrode (AgQRE).
However, in order to ensure maximal stability of the reference
potential in those experiments in which high concentrations of
lead anionic species were generated, a regular Ag/Ag+ reference
electrode was used. This reference electrode was prepared by
electrochemical oxidation of the Ag wire in situ to produce some
Ag+ cations in the reference electrode compartment. Such a
procedure prevented a shift of the silver electrode potential even
if some amount of Pb94- diffused through the frits and entered
the reference electrode compartment during the long-lasting
experiments. Occasionally the AgQRE was placed in a small
fritted compartment that was placed in the main reference
electrode compartment. The potentials reported in the present
paper are referenced versus the solvated electron potential. The
potential used in this study is the oxidation peak potential of
solvated electrons generated voltammetrically on a 25 µm Pt
UME. The solvated electron was generated at a scan rate of
50mV/s with the reversal potential at which the reduction current
reached approximately 12 µA. The reference potential deter-
mined in this way was virtually ohmic drop free. Sometimes a
potential which corresponds to a certain cathodic current density
4-
to include heteroatomic species, such as for example Sb2Te5
,
3- 14
Sb6Te9
,
4- 13 or Cu4SbTe12
.
Products of reduction of cathodes
made of mercury15-23 and other metals24-28 in solutions of
quaternary ammonium salts in some organic solvents were
suggested to be alkylammonium salts of the respective homo-
atomic Zintl ions.15,28 Contemporary, well-controlled, electro-
chemical generation methods were applied to produce hetero-
4-
atomic PbxSn9-x species29 as well as alkylammonium or
alkylphosphonium salts of some heteroatomic30-33 Zintl ions.
Detailed electrochemical studies of reduction processes of some
metal cathodes in certain organic solvents containing organic
background electrolytes are also available.17-28 These studies17-28
generally show complicated character of the reduction processes
involving the electrode material. Quite unexpectedly, no
detailed studies have been performed on the electrochemical
behavior of homoatomic Zintl ions in the presence of alkali
metal salts. Rather simpler behavior may be expected for such
systems, and we believe that a better understanding of the
cathodic reduction (corrosion) of metals could be achieved with
the aid of such studies. Such studies will also provide a deeper
insight into some unusual chemical reactions of Zintl ions,34-37
which are certainly electrochemical in nature. Lead, for which
the Pb94-, Pb74-, and Pb52- Zintl ions are known, was the first
element which was selected, because of the relative ease of
production of lead electrodes in variety of sizes and shapes.
Liquid ammonia was selected as a solvent, because simple and
very efficient purification methods of that solvent for electro-
chemical purposes are known.38,39 Also, the electric permittivity
of liquid ammonia is close to 25, which makes this solvent more
X Abstract published in AdVance ACS Abstracts, March 15, 1997.
S1089-5647(96)03073-8 CCC: $14.00 © 1997 American Chemical Society

2868 J. Phys. Chem. B, Vol. 101, No. 15, 1997
Chlistunoff and Lagowski
is used as a reference potential in liquid ammonia,38 a condition
that moves us to also quote the corresponding value for the
reference potential. The reference potential corresponds to the
reduction current density of 22.6 mA/cm2. The experiments
were performed using a EG&G PARC 273A potentiostat.
Unless otherwise stated, a positive feedback was used as a
method of iR drop compensation.
In a typical experiment the electrochemical cell was loaded
with KI Reagent, ACS (MCB, Cincinnati, OH) and connected
to a vacuum line. It was dried at 120-130 °C under vacuum
for 2 or 3 days. Before the electrodes were placed in the cell,
a fresh surface of Pb was exposed either by a brief grinding
with a triangular file (epoxy) or by cutting off a tip of the
electrode with a razor blade (PVC). Grinding of the Pb/epoxy
electrodes required some skill to prevent the lead from spreading
all over the epoxy. The freshly prepared electrodes were loaded
into the cell, which was then evacuated under vacuum at 110
°C for 1-2 h. The reason for not heating the electrodes for a
long time was a possible damage to their surface or shell. The
Pb/PVC electrodes could not be heated above 120 °C and/or
for long periods of time, since such a treatment resulted in
irreversible loss of PVC elasticity and consequently in a poor
Pb-PVC seal. The lead surface of the Pb/epoxy electrodes,
on the other hand, was easily oxidized at elevated temperatures
by the moisture accumulated in the epoxy. Because of that
possibility, the Pb/epoxy electrodes were predried under vacuum
in a separate vessel and kept under vacuum until they were used
in an experiment. Ammonia (anhydrous grade, Matheson) was
condensed under vacuum in a trap containing metallic sodium
with liquid nitrogen. After condensation three freeze-pump-
thaw cycles were applied to remove any gaseous impurities left
after purification with sodium. Finally, the purified ammonia
was vacuum-distilled into the electrochemical cell. Ammonia
solutions obtained by this process were usually sufficiently pure
for electrochemical experiments as demonstrated by cyclic
voltammogram on a Pt disk electrode. In very rare cases,
however, some reducible impurities were still present as
manifested by the lowered reoxidation current of solvated
electrons generated at the cathodic potential limit. In such cases
small concentrations of solvated electrons were generated at a
constant potential in order to reduce the impurities. This
procedure was repeated as many times as necessary, i.e., until
the light blue color of a diluted e- solution could be seen for
more than 5 min.
Figure 1. Number of electrons associated with the removal of a single
lead atom during the cathodic reduction of lead plotted against the
reduction potential.
to iR drop effects. During a similar study with RbBr and CsI
solutions,40 in which formation of a solid compound occurs at
a lower overpotential, solid products of a preparative electrolysis
were easily detected. Stoichiometry of the dissolved products
was determined using a gravimetric method. The lead cathode
was weighed before and after preparative electrolysis, and the
number of electrons (n) associated with removal of a single lead
atom from the electrode was calculated from the electrolysis
charge and the electrode weight loss. The plot of n versus the
electrolysis potential, shown in Figure 1, suggests that the dark
4-
4-
colored species generated was Pb9 (Pb0.44(4)-). That Pb9
was the only reduction product was further confirmed by
voltammetric experiments described below. The Pb94- solution
could be reoxidized at 1.254 V (see below) with an efficiency
exceeding 95%, thus suggesting that the oxidation product was
metallic lead. The deposit was black, spongelike, and did not
adhere well to the electrode surface. Identical behavior was
observed when PbI2 solution was reduced under similar condi-
tions and is rather typical for electrochemical deposition of lead
both in liquid ammonia41 and in water.42
It has to be noted that a long electrolysis can result in
accumulation of NI3 on the Pt anode. Since NI3 is explosive
when dry, care has to be exercised when removing the counter
electrode from the cell and when cleaning it.
4-
Cyclic Voltammetry. The electrochemical behavior of Pb9
was studied using both pure KI solutions and solutions contain-
ing Pb94- produced by preparative electrolysis using a large lead
cathode (see above). All the data described below were obtained
after electrochemical activation of the lead surface. The
activation procedure was not always necessary, but ocassionally
the surface of the electrode was partially oxidized. The
activation consisted of a potential scan in the cathodic potential
region until the steady state cyclic voltammogram was obtained.
The potential range applied was that in which no solid reduction
products were generated (see below). For 0.1 M KI solutions
potentials between approximately +0.5 and +0.3 V fulfilled
this requirement. Usually, fewer than 10 cycles at a scan rate
of 50 mV/s were necessary to activate the surface. Once
activated, the surface remained active during experiments.
At relatively low cathodic potentials (E) the Pb reduction
current in 0.1 M KI is independent of the scan rate and follows
approximately exponential increase with E (Figure 2) as
expected for a completely irreversible process. A single
reoxidation peak can be seen at around 0.6 V. No other products
could be detected for scan rates ranging from 20 mV/s to 20
V/s. However, the reoxidation currents recorded for scan rates
V e 0.1 V/s were lower than expected (curve 2 in Figure 2 and
Figure 3), suggesting instability of the product. However,
All the experiments were performed at the dry ice/2-propanol
bath temperature (-70 °C).
Results
4-
Preparation of Pb9 Solutions in Liquid Ammonia.
Preparative electrolysis in 0.1 M KI solutions was performed
at selected cathodic potentials more positive than the solvated
electron potential. The potentials were not corrected for iR drop
and therefore apply only to the specific cell geometry. Com-
pensation for iR in the preparative experiments case is of rather
limited usability, because of the large surface area of the
electrode. Some parts of the electrode would remain under-
compensated, some other would be overcompensated, and
comparison of the potentials with the corresponding voltam-
metric data would still be doubtful. The reduction produced a
very dark solution which initially appeared violet and then
brownish red. The lead electrode had a shiny appearance after
electrolysis, and a characteristic corrosion pattern could be seen
on its surface irrespective of the reduction potential applied.
No solid deposit formation was detected in contrast to volta-
mmetric data (see below). This discrepancy is most likely due

4-
Lead Zintl Ion Pb9
J. Phys. Chem. B, Vol. 101, No. 15, 1997 2869
1.0 M KI solutions. The natural convection effect seems to be
especially strong43 in the present case, and the most probable
reason for that is the high molecular weight of the metal clusters
generated.
When the cathodic scan limit is extended up to 0 V, the
reduction current does not follow exponential increase. At low
scan rates (V e 0.05 V/s) a peak current followed by some
current increase could be seen in the cathodic scan (Figure 4a).
The cathodic current in the reverse scan exhibits hysteresis
behavior and suggests chemical modification of the surface by
an intermetallic compound of lead and potassium. Our other
studies40 indicate that the molar ratio of lead to potassium in
the compound is close to one. The modified surface is oxidized
in a broad irregular surface peak (a2) between 0.5 and 0.9 V,
which overlaps with the oxidation peak (a1) of the dissolved
species as described above. A relatively small oxidation current
is also seen at the anodic potentials preceding the broad peak
(0.4 V < E < 0.6 V). This current transforms into a diffuse
peak (a3) at approximately 0.5 V at higher scan rates (V g 0.1
V/s). At the same time the broad peak (a2), centered around
0.7 V, disappears. At still higher scan rates (V g 0.2 V/s) the
peaks a1 and a2 merge, or peak a2 disappears and a small anodic
peak (a4) appears at around 0.35 V (V g 0.5 V/s). For scan
rates in the range 0.5 V/s-5 V/s the peaks a1 and a4 are the
only well-resolved peaks. The ratio of their heights does not
seem to vary with the scan rate. Voltammetric behavior of Pb
in 1.0 M KI was essentially similar; however, some differences
were found. As many as three cathodic peaks could be seen at
low scan rates for 1.0 M solutions (Figure 4b). The anodic
peak a4 was much better resolved, and larger amounts of the
solid compound were formed (Figure 4b). The last two effects
are probably associated with generally higher reduction currents
measured in 1.0 M KI solution.
Figure 2. Cyclic voltammetric curves recorded using a 1 mm Pb disk
electrode in 0.1 M KI solution in liquid ammonia at -70 °C without
4-
(curves 1 and 2) and with 0.4 mM concentration of Pb9 (curve 3).
Starting potential: +0.494 V (curves 1 and 2); 0.526 V, i.e., open-
circuit potential (curve 3). Scan rate (V/s): 0.05 (curve 2); 0.2 (curves
1 and 3). Direction of the forward scan: cathodic (curves 1 and 2);
anodic (curve 3).
Voltammograms of the solutions prepared by the preparative
electrolysis exhibited only a single oxidation peak whose
position and slope matched well with the same parameters for
peak a1 (Figure 2). This fact strongly suggests that a single
dissolved species is generated during preparative electrolysis,
since, due to structural differences and kinetic factors, the known
lead Zintl ions (Pb94-, Pb74-, and Pb52-) are expected to oxidize
at different potentials. Therefore, in view of the coulometric
results, the peak a1 is attributed to the oxidation of Pb94-. The
4-
electrochemically irreversible oxidation of Pb9 on a lead
electrode is diffusion controlled (Figure 5) in both 0.1 and 1.0
M KI solution. We found, however, that recording a large
number of Pb94- oxidation curves resulted in some increase of
the oxidation current, especially when higher concentrations of
Figure 3. Ratio of the oxidation charge of Pb94- to the reduction charge
of 1 mm lead disk electrode (Pb9 generation) in 0.1 M KI solution
under voltammetric conditions plotted versus the square root of the
scan rate. Starting potential +0.494 V. Cathodic reversal potential
+0.284 V. Anodic reversal potential +1.054 V.
4-
4-
Pb9 were employed. This effect suggests that the active
electrode surface area increases during subsequent oxidation
scans. Examination of the electrodes after performing multiple
oxidations revealed the presence of the fine black Pb deposit
on the surface (see also coulometric data). When the electrode
potential was scanned cathodically up to +0.3 V at a scan rate
of 50 mV/s, the black lead deposit dissolved easily, exposing a
fresh, shiny bulk lead surface. Therefore, such a pretreatment
(in fact identical with the activation procedure described above)
coulometric data (see above) confirmed stability of the lead
reduction product. The decrease in the reoxidation charge is
then interpreted as being due to a natural convection. The effect
occurs at relatively slow scan rates (long reduction times)43 when
significant amounts of the product of an electrode reaction are
generated and is caused by the density gradients in the vicinity
of the electrode surface. The density gradient results in
convection which enables faster transport of the product away
from the electrode surface. At the low scan rates, at which the
lowered reoxidation charges were seen (Figures 2 and 3), the
reduction process was accompanied by dark streamers going
downward from the Pb disk surface, in accordance with our
interpretation of the lowered reoxidation charges. As expected,
the effect occurred to a lesser extent in more viscous and denser
4-
of the lead disk was performed occasionally during Pb9
oxidation studies to ensure reproducible surface area of the
electrode. As would be expected, the presence of the fine lead
deposit on the electrode surface also leads to an increase in the
lead reduction currents (Figure 6). The effect is especially
significant when the lead is reduced immediately after oxidation
of the solid formed at the potentials close to solvated electron
generation (Figure 7). A similar phenomenon was reported for
Sb cathodes reduced in dimethylformamide solutions of qua-

2870 J. Phys. Chem. B, Vol. 101, No. 15, 1997
Chlistunoff and Lagowski
4-
Figure 5. Plot of the peak current for oxidation of 0.4 mM Pb9 in
0.1 M KI solution on a 1 mm Pb disk electrode versus the square root
of the scan rate. The starting potential was the open-circuit potential
of the lead electrode in the solution studied.
4-
Figure 6. Effect of oxidation of large quantities of Pb9 on the 1
mm Pb disk reduction currents in 0.1 M KI solution. Curve 1: Starting
potential +0.284 V for 20 s, then the potential scanned anodically up
to +1.054 V and cathodically to +0.494 V. Curves 2 and 3: two
subsequent scans between +0.494 and +0.284 V recorded immediately
after recording curve 1. Scan rate 0.5 V/s.
Figure 4. (a, top) Cyclic voltammograms recorded using 1 mm lead
disk electrode in 0.1 M KI solution in liquid ammonia at -70 °C.
Starting potential +0.494 V. Cathodic reversal potential -0.026 V.
Anodic reversal potential +1.054 V. Scan rate (V/s): 0.05 (1); 0.1
(2); 0.5 (3); 2.0 (4). (b, bottom) Cyclic voltammograms recorded using
1 mm lead disk electrode in 1.0 M KI solution in liquid ammonia at
-70 °C. Starting potential +0.524 V. Cathodic reversal potential (V):
+0.344 (1); +0.164 (2-4); -0.026 (5). Scan rate (V/s): 0.02 (5); 0.1
(1); 0.2 (4); 0.5 (3); 1.0 (2). Anodic reversal potential (V): +1.024
(1-3); +1.124 (4, 5).
Figure 7. Two subsequent cathodic and anodic scans (0.02 V/s) on a
1 mm Pb disk electrode between +0.494 and +0.284 V in 0.1 M KI
solution recorded immediately after voltammetric reduction of the
electrode between +0.494 V and 0 V and reoxidation of the products
4-
(Pb9 and a solid phase, see text and curve 1 in Figure 3). The peak
seen in the first cathodic scan originates from reduction of finely divided
lead produced upon oxidation of the solid compound.
ternary ammonium salts.28 The large magnitude of the effect
is easily understood if the surface compound formed is an
intermetallic compound of lead and potassium. Deintercalation
of K upon oxidation of the compound would leave a layer of a
very porous, probably amorphous lead characterized by an
exceptionally large surface area. After a few potential cycles
between 0.5 and 0.3 V (0.1 M KI) the original reduction currents
can be restored (Figures 6 and 7). The cathodic dissolution of
the porous lead deposit was employed to clean the electrode
surface after the oxidation of the surface compound.

4-
Lead Zintl Ion Pb9
J. Phys. Chem. B, Vol. 101, No. 15, 1997 2871
4-
The peak potential of Pb9 oxidation at 50 mV/s is 0.57 V
in 0.1 M KI and 0.61 V in 1.0 M KI. The difference between
the peak potential and the potential at the half-height is 18 and
16 mV, respectively, which corresponds to an anodic transfer
coefficient ân close to 2 and suggests that the structure of the
transition state is intermediate between the structures of the
metallic lead and the Pb94- in solution. The peak a4, which is
most probably a surface peak, corresponds to oxidation of some
transition product of the electrolysis. Our other studies40 show
that the peak a3 corresponds to oxidation of an intermetallic
compound other than the one oxidized in peak a2. The detailed
analysis of the solids formed during the reduction of a lead
cathode in liquid ammonia solutions of KI, RbBr, and CsI will
be presented in a subsequent paper.40
Chronocoulometry and Chronoamperometry. Determi-
nation of the Diffusion Coefficient of Pb94-. Chronocoulom-
etry is known to be a reliable method for determining the dif-
fusion coefficient of the electroactive species when the number
of electrons transferred and concentration of the electroactive
species are known. Alternatively, the concentration of the elec-
troactive species can be easily determined chronocoulometrically
if the diffusion coefficient is known. The actual volume of the
solution in our small three-compartment cell could only be
estimated, because of the relatively large void volume associated
with presence of the electrodes as well as partial diffusion of
the reduction products into the frits which separated the cell
4-
compartments. As a result, neither the concentration of Pb9
nor its diffusion coefficient could be determined precisely using
just chronocoulometry. Therefore, we tried to apply a method,
suggested recently by Bard and co-workers,44 which is based
on a measurement of an ultramicroelectrode current vs time
transient at a potential where the current is limited solely by
the rate of diffusion. This chronoamperometric method is
accurate and allows for the determination of the diffusion
coefficient without knowing the number of electrons transferred
or concentration of the electroactive species. In a simplest
approach the diffusion of a species toward an ultramicroelec-
trode can be regarded as a sum of a linear (time dependent)
and convergent or spherical (time independent) components. The
linear component is proportional to the reciprocal square root
of time, while the spherical component is a constant which
depends on the electrode radius. At the very beginning of
electrolysis the current is mainly determined by the linear
component (1/t1/2 f ∞). After sufficiently long time the
spherical component is predominant (1/t1/2 f 0), and the current
is referred to as a steady state current. The current normalized
with respect to the steady state current is then given by44
Figure 8. (a) Plot of a current versus inverse square root of time for
a potentiostatic oxidation of 0.4 mM Pb9 in 0.1 M KI solution at
4-
+1.054 V on a 25 µm Pt ultramicroelectrode. Initial potential +0.494
V. Anodic pulse duration 1 s. Current measured every 1 ms. Nonlinear
section of the plot at highest 1/t-1/2 (first 3 ms) is believed to originate
from the electrode surface charging and a delayed beginning of the
lead deposition due to a slow nucleation process. (b) Chronocoulometric
charge vs square root of time plot obtained for the potentiostatic
4-
oxidation of 0.4 mM Pb9 in 0.1 M KI at +1.054 V on a 1 mm Pb
disk electrode. Initial potential corresponds to the open-circuit potential
of the disk in the solution studied. Anodic pulse duration 0.1 s. The
charge measured every 0.1 ms.
a simple reliability test based on comparison of the slope of
the i(t) vs t-1/2 plot with the slope of a charge vs t1/2 plot for
chronocoulometric experiment with the large lead disk in the
same solution. The ratio of the chronocoulometric to the
chronoamperometric slope should be twice the ratio of surface
areas of the electrodes used in both experiments. Since
preparation of a lead ultramicroelectrode (UME) of a specific
geometry is not an easy task, we decided to employ 25 µm Pt
UMEs for the Pb94- diffusion coefficient determination. In this
i(t)/iss ) (π1/2/4)r(Dt)-1/2 + 1
(1)
where D is the diffusion coefficient and r is the electrode radius.
The diffusion coefficient can be determined from the slope of
the i(t)/iss vs t-1/2 plot. This method requires that the steady
state current and the ultramicroelectrode radius are known.
However, determination of the steady state current in the present
case is problematic, because of the accumulation of the metallic
lead on the electrode surface (see below). The same problem
remains for more accurate methods44 where nonlinear equations
are used to describe i(t)/iss. Therefore, we decided to use the
linear approximation (eq 1) with an assumption that the steady
state current can be determined from the intercept of the i(t) vs
t-1/2 plot with the current axis. Such a method is expected to
minimize the effect of the deposited metal on the electrode
radius (see below), since the steady state current is obtained by
extrapolation from the short time region, where the size of the
electrode has not been altered yet. The method also allows for
4-
case, however, the Pb9 oxidation on Pt had to be studied in
more detail to establish the possibility as well as conditions of
using Pt UME for such a purpose. Almost instantaneous
nucleation45 of lead on the whole Pt surface followed by a
diffusion-controlled oxidation was found during potentiostatic
4-
oxidation of 0.7 mM Pb9 solution on a 0.5 mm Pt disk at a
high anodic potential of +1.254 V. However, the very next
transient recorded with the same electrode exhibited slightly
higher currents as well as current spikes even though the
cathodic cleaning (see above) was employed. Experiments with
a more concentrated Pb94- solution and Pt wire electrode, which
could be easily observed during experiment, revealed that part
of the lead deposit looses its electrical contact with Pt during
the cathodic cleaning, but it is still mechanically attached to it.
During subsequent oxidation of Pb94-, the electrical contact can

2872 J. Phys. Chem. B, Vol. 101, No. 15, 1997
Chlistunoff and Lagowski
4-
be reestablished which results in generally higher oxidation
currents and occasionally current spikes. Therefore, Pt elec-
trodes could only be used to record a single transient. Similar
effects were expected to be even more pronounced with the
UMEs because of generally larger amounts of Pb deposited per
unit surface area of the electrode due to the increased transport
rate in the spherical diffusion field. Indeed, a progressive
increase of the oxidation current has been observed when a
single microelectrode was used to perform multiple oxidations.
However, the very first transients recorded for relatively less
The fact that Pb9 was detected as the only stable product
of the reduction suggests that the other possible anions, i.e.,
Pb7 and Pb52-, are rather unstable under our experimental
4-
conditions. This finding can be associated with the use of an
excess of the KI electrolyte in the present study and possible
ion association effects.1 The Pb74- was produced9 by a titration
of sodium solution in liquid ammonia with PbI2. In such an
experiment, not only was the countercation different than in
the present study but also its concentration was comparable with
the Zintl ion concentration. The Pb52-, on the other hand, was
detected in the solid state48 rather than in solution.
4-
concentrated Pb9 solutions produced results that were con-
cordant with the chronocoulometric data obtained with a 1 mm
Pb disks. The slope of the chronoamperometric (i vs t-1/2) plot
obtained in such a case was typically 3000-3300 lower than
the slope for corresponding chronocoulometric (Q vs t1/2) plot,
while the expected ratio was 1:3200. Such results (Figure 8)
As already mentioned, the electrochemical reduction of lead,
which produces Pb94-, is a complicated reaction. No detailed
mechanism of that reaction can be deduced from the present
data. However, our studies including other electrolytes, e.g.,
NaI, NaClO4, RbBr, and CsI,40,46 are in progress, and some
important conclusions have already been reached. These studies
will be published soon.
4-
were used to determine the diffusion coefficient of Pb9 in
0.1 M KI solution. The average value obtained from six
independent experiments is (4.13 ( 0.23) × 10-6 cm2/s. The
4-
Acknowledgment. The support of this research by the Welch
Foundation (Grant F-0081) is gratefully acknowledged. Thanks
are also due to Mr. Steve Lake for discussions.
Pb9 concentrations in freshly prepared solutions which were
calculated using the above diffusion coefficient as well as
chronocoulometric data agreed within 10% with the correspond-
ing values calculated from the electrolysis charge and ap-
proximate volume of the solution. In a few cases, the con-
centrations were lower than the estimated ones, most probably
because of partial oxidation of the Pb94- solution to zerovalent
lead by products generated in the anodic compartment which
may have diffused into the cathodic compartment during long
experiments. One should note that no decomposition can be
detected by visual examination of the solution in the cell, since
the solution appears almost black even at submillimolar con-
centrations. Consequently, chronocoulometry on a lead disk
electrode can be recommended as an accurate method for
determination of Pb94- concentration in KI solutions, especially
in cases when partial decomposition has occurred.
References and Notes
(1) Corbett, J. D. Chem. ReV. 1985, 85, 383 and references therein.
(2) Johannis, A. C. R. Hebd. Seances Acad.Sci. 1891, 113, 759 (quoted
in ref 1).
(3) Johannis, A. C. R. Hebd. Seances Acad. Sci. 1892, 114, 587 (quoted
in ref 1).
(4) Kraus, C. A. J. Am. Chem. Soc. 1907, 29, 1557.
(5) Smyth, F. H. J. Am. Chem. Soc. 1917, 39, 1299.
(6) Peck, E. B. J. Am. Chem. Soc. 1918, 40, 335.
(7) Kraus, C. A. J. Am. Chem. Soc., 1922, 44, 1216.
(8) Kraus, C. A. Trans. Am. Electrochem. Soc. 1924, 45, 175 (quoted
in ref 1).
(9) Zintl, E.; Goubeau, J.; Dullenkopf, W. Z. Phys. Chem., Abt. A 1931,
154, 1.
(10) Zintl, E.; Harder, A. Z. Phys. Chem., Abt. A 1931, 154, 47.
(11) Zintl, E.; Dullenkopf, W. Z. Phys. Chem. Abt. B 1932, 16, 183.
(12) Zintl, E.; Kaiser, H. Z. Anorg. Allg. Chem. 1933, 211, 113.
(13) Warren, C. J.; Dhingra, S. S.; Ho, D. M.; Haushalter, R. C.;
Bocarsly, A. B. Inorg. Chem. 1994, 33, 2709.
(14) Dhingra, S. S.; Haushalter, R. C. J. Am. Chem. Soc. 1994, 116,
3651.
(15) Garcia, E.; Cowley, A. H.; Bard, A. J. J. Am. Chem. Soc. 1986,
108, 6082.
(16) Littlehailes, J. D.; Woodhall, B. J. Chem. Commun. 1967, 665.
(17) Kariv-Miller, E.; Nanjundiah, C. J. Electroanal. Chem. 1983, 147,
319.
Discussion
To our knowledge, this is the first study of electrochemical
behavior of homoatomic Zintl ions. Among the systems studied
by us so far, including lead40,46 and bismuth47 Zintl ions in liquid
ammonia solutions of alkali metal salts, the present system is
the simplest one. Even though, only the oxidation reaction of
4-
Pb9 in the KI solution can be regarded as a simple process.
(18) Kariv-Miller, E.; Nanjundiah, C.; Eaton, J.; Swenson, K. E. J.
Electroanal. Chem. 1984, 167, 141.
(19) Kariv-Miller, E.; Svetlicic, V. J. Electroanal. Chem. 1985, 205,
319.
(20) Kariv-Miller, E.; Andruzzi, R. J. Electroanal. Chem. 1985, 187,
175.
(21) Svetlicic, V.; Kariv-Miller, E. J. Electroanal. Chem. 1986, 209,
91.
(22) Ryan, C. M.; Svetlicic, V.; Kariv-Miller, E. J. Electroanal. Chem.
1987, 219, 247.
(23) Ryan, C. M.; Svetlicic, V.; Kariv-Miller, E. J. Chem. Soc., Faraday
Trans. I 1988, 84, 4023.
In addition, the simple character of the reaction is limited to
certain experimental conditions. However, the fact that Pb9
4-
can be oxidized both in a simple way and with accompanying
side effects allowed us to determine the nature of the side effects
and to define the conditions required to avoid or minimize them.
The effects associated with fine crystalline morphology of the
4-
lead deposits formed upon oxidation of Pb9 and lead-
potassium intermetallic compounds affect both the lead reduction
4-
and Pb9 oxidation processes and are especially dangerous,
since the results distorted by these effects quite often look
normal. Understanding these effects has already made the
interpretation of our other results46,47 easier. It also allows us
to determine correctly the diffusion coefficient of Pb94- in 0.1
M KI solution and apply the obtained value in a simple
chronocoulometric method of Pb94- determination. The method
is very simple and much more convenient than other methods
which could be applied for analysis of Pb94- solutions in liquid
ammonia. For instance, a spectrophotometric method using the
visible region may be inferred to be useful, because Pb94- is a
colored species. Instead, its use is rather limited in the
millimolar concentration range because the light absorption by
(24) Kariv-Miller, E.; Lawin, P. B.; Vajtner, Z. J. Electroanal.Chem.
1985, 195, 435.
(25) Kariv-Miller, E.; Lawin, P. B. J. Electroanal. Chem. 1988, 247,
345.
(26) Lawin, P. B.; Svetlicic, V.; Kariv-Miller, E. J. Electroanal. Chem.
1989, 258, 357.
(27) Svetlicic, V.; Lawin, P. B.; Kariv-Miller, E. J. Electroanal. Chem.
1990, 284, 185.
(28) Fidler, M. M.; Svetlicic, V.; Kariv-Miller, E. J. Electroanal. Chem.
1993, 360, 221.
(29) Pons, S. B.; Santure, D. J.; Taylor, R. C.; Rudolph, R. W.
Electrochim. Acta 1981, 26, 365.
(30) Warren, C. J.; Ho, D. M.; Haushalter, R. C.; Bocarsly, A. B. Angew.
Chem., Int. Ed. Engl. 1993, 32, 1646.
(31) Warren, C. J.; Ho, D. M.; Bocarsly, A. B.; Haushalter, R. C. J.
Am. Chem. Soc. 1993, 115, 6416.
4-
Pb9 is very strong.

4-
Lead Zintl Ion Pb9
J. Phys. Chem. B, Vol. 101, No. 15, 1997 2873
(32) Warren, C. J.; Ho, D. M.; Haushalter, R. C.; Bocarsly, A. B. J.
Chem. Soc., Chem. Commun. 1994, 361.
(33) Warren, C. J.; Haushalter, R. C.; Bocarsly, A. B. Chem. Mater.
1994, 6, 780.
(34) Haushalter, R. C. Angew. Chem. Suppl. 1983, 766.
(35) Haushalter, R. C.; Krause, L. J. Thin Solid Films 1983, 102, 161.
(36) Treacy, M. M. J.; Haushalter, R. C.; Rice, S. B. Ultramicroscopy
1987, 23, 135.
(37) Haushalter, R. C.; Treacy, M. M. J.; Rice, S. B. Angew. Chem.
1987, 99, 1172.
(41) Booth, H. S.; Merlub-Sobel, M. J. Phys. Chem. 1931, 35, 3303.
(42) Sharpe T. F. In Encyclopedia of Electrochemistry of the Elements;
Bard, A. J., Ed.; Marcel Dekker: New York, 1973; Chapter I-5.
(43) Galus, Z. Fundamentals of Electrochemical Analysis, 2nd revised
ed.; Ellis Horwood: New York, 1994; Chapter 5, p 93.
(44) Denault, G.; Mirkin, M. V.; Bard, A. J. J. Electroanal. Chem. 1991,
308, 27.
(45) Vargas, T.; Varma R. In Techniques for Characterization of
Electrodes and Electrochemical Processes; Varma, R., Selman, J. R., Eds.;
John Wiley & Sons: New York, 1991; Chapter 15.
(38) Crooks, R. M.; Bard, A. J. J. Phys. Chem. 1987, 91, 1274.
(39) Gill, J. B.; Hall, M.; Taylor, N. Electrochim. Acta 1988, 33, 89.
(40) Chlistunoff, J. B.; Lagowski, J. J. Manuscript in preparation.
(46) Chlistunoff, J. B.; Lagowski, J. J. Unpublished results.
(47) Lake, S. M.; Lagowski, J. J. unpublished results.
(48) Edwards, P.; Corbett, D. J. Inorg. Chem. 1977, 16, 903.